IMPLEMENTATION AND PERFORMANCE ANALYSIS OF COGNITIVE RADIO WITH FREQUENCY UPDATING ALGORITHM ON SOFTWARE DEFINED RADIO PLATFORM by Jaskaran S. Phull
Abstract
The field of wireless communication is ever growing and over the last 6 decades, the wireless standards have revolutionized the field of communication. Wireless communication is being used in all communication standards. For some of the advanced standards, the performance depends upon high gain. And for achieving high gain, the spectrum bandwidth is necessary. But day by day bandwidth scarcity is becoming a major concern for the upcoming wireless technologies. For this various techniques based on artificial intelligence has been designed. The basic intelligent radio called as cognitive radio has been designed which works on the basic principle of spectrum sensing and detecting the free frequency for transmission of secondary user which is basically unlicensed user. Conventionally, sincere efforts have been made to develop the algorithms or techniques for designing of the cognitive radio on MATLAB platform and also on other software’s intend for radio designing. In this work, an efficient technique has been developed to design cognitive radio based on SDR platform. On this platform, GNU Radio Companion has been used as the software part for signal processing and Universal Software Radio Peripheral is used for real time signal transmissions for designing of cognitive radio. The spectrum sensing energy threshold detection technique is used for this. In this work , a carrier frequency updating algorithm has been added to assess the performance of the proposed technique. This algorithm updates the frequency of secondary user in Linux terminal window. This makes the radio more reliable and efficient. Further the effect of noise on the secondary transmitter and spectrum sensor is also analysed using the SDR approach. The increased noise levels in the environment leads to frequency updating of secondary user even in absence of primary user. It is observed from the analysis that for every 10dB rise in Gaussian Noise, the bit error rate of secondary transmitter and spectrum sensor cause increment of 19.59% and 29.39% respectively. It is found that spectrum sensor is more easily prone to noise and the Gaussian noise causes the degradation of the system performance. It is further noted that the spectrum sensor should be programmed carefully to avoid the effect of Gaussian noise. From the above analysis, it is observed that best range of spectrum sensor under Gaussian noise is 0 to 0.1 dB to obtain minimum bit error rate. In future more work can be done using high range USRP as well as IP (Internet Protocol) based USRP which will make our radio more smarter and efficient. Even interfacing the GNU radio with MATLAB can prove better results without using the Python coding.
